<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
ai-light-dance_singing_ft_wav2vec2-large-lv60-v2
This model is a fine-tuned version of gary109/ai-light-dance_singing_ft_wav2vec2-large-lv60 on the ../AI_LIGHT_DANCE.PY - ONSET-SINGING dataset. It achieves the following results on the evaluation set:
- Loss: 0.4285
- Wer: 0.1858
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 3e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 10.0
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Wer |
---|---|---|---|---|
0.2775 | 1.0 | 1106 | 0.4372 | 0.2117 |
0.2154 | 2.0 | 2212 | 0.4474 | 0.2044 |
0.2023 | 3.0 | 3318 | 0.4372 | 0.1920 |
0.186 | 4.0 | 4424 | 0.4285 | 0.1858 |
0.1856 | 5.0 | 5530 | 0.4589 | 0.1826 |
0.1537 | 6.0 | 6636 | 0.4658 | 0.1774 |
0.1337 | 7.0 | 7742 | 0.4769 | 0.1744 |
0.108 | 8.0 | 8848 | 0.4604 | 0.1724 |
0.1593 | 9.0 | 9954 | 0.4731 | 0.1694 |
0.0904 | 10.0 | 11060 | 0.4843 | 0.1683 |
Framework versions
- Transformers 4.20.0.dev0
- Pytorch 1.11.0+cu102
- Datasets 2.2.2.dev0
- Tokenizers 0.12.1